| Abstract: | Dementia is a growing health and economic concern around the world, with the absolute growth of its numbers estimated to be highest in East Asia (Hogervorst et al., 2021). The situation in Taiwan, which is the third most aged country in Asia and will enter the super-aged society in 2025, should be critical. A recent study using Taiwan's National Health Insurance (NHI) claims data in this issue found that the incidence of dementia increased slightly but the number of cases increased dramatically, and the annual mortality in the sub-cohort of incident dementia patients was approximately 12 % during 5-year follow-up (Huang et al., 2024). This study was designed to use the full-population NHI data with the more granular ICD-10 disease coding system, which has been implemented in NHI since 2016 to characterize dementia patients in Taiwan. An advantage of the NHI data source is the low membership turnover, which is particularly important for long term follow-up studies. Another similar study using data from the National Health Insurance Service (NHIS) in Republic of Korea also showed increasing trends in the incidences of all-cause dementia and Alzheimer's disease (AD) dementia (Hwangbo et al., 2023). However, a number of previous research findings from high-income western countries revealed that the incidence rates of dementia may have declined in the last 25 years (Ding et al., 2020; Farina et al., 2022; Wolters et al., 2020), although some results were inconsistent (Alzheimer's Association, 2023; Chen et al., 2023). When we compare these conflicting findings from different study settings, we have to be careful because it is not clear whether these encouraging declines in incidence occurs in all racial and ethnic groups (Power et al., 2021). Both field epidemiologic survey and claims data have their pros and cons. For example, the lack of “gold standard” such as rigorous cohort evaluations in claims data and the efforts on validating the diagnosis in claims data are scare (Grodstein et al., 2022). Routinely collected administrative claims data may confer a potential barrier as they can be linked to baseline data even a person has been lost to follow-up. In addition, a report stated that measuring dementia incidence in a population or cohort study is difficult because dementia remains undiagnosed in the early stages and persons with dementia may be disproportionately missed from surveys or follow-up due to their cognitive deficits (Welberry et al., 2020). |
